EN

Tensile Stress, Elongation & Safety Factor

A part pulled at both ends — a cable, a tie rod, a bolt, a test specimen — carries a normal stress σ equal to the force divided by the cross-section. This calculator returns the stress, the strain and the elongation from Hooke's law, then puts the result in perspective against the material's yield strength: the safety factor tells you in a single number whether the part holds, and by what margin.

The whole strength of materials fits in this comparison: the applied stress (which depends only on geometry and load) against the allowable stress (which depends only on the material).

Load, geometry and material

Load and geometry

Smallest cross-section (at a thread: tensile stress area As).

Material

E and Re fill in automatically (except "Custom").

Used only when "Custom" is selected.

Used only when "Custom" is selected.

Terms marked with an asterisk are defined in the glossary.

Uniform tension: two collinear forces and one normal stress

A bar pulled at both ends is the simplest loading strength of materials treats. Two equal forces, opposite in sense, act on the same axis. The cross-section answers with a uniform normal stress, perpendicular to the cut: σ = F / A. Bending does not appear, torsion does not appear, and shear does not enter. The response is extension along the axis.

σ = F / A  ·  ε = σ / E  ·  ΔL = ε × L₀  ·  s = Re / σ

While σ stays below the yield strength* Re, the bar lengthens in proportion to the load and recovers its original length once the load is removed. That is the elastic range. Past Re the bar is permanently longer. The calculation on this page stays inside that first regime: Hooke's law*, a straight stress–strain line, and a factor of safety* that still refers to a recoverable elongation.

Uniform stress is an assumption. It fits a prismatic bar, a centred pull, and a section with no hole, no thread and no weld toe. Real fittings seldom match that picture; the sections further down return to the details that break it. In the formula itself, A is the load-bearing area at the thinnest station — the station that reaches yield first.

Hooke's law, and why stiffness is not strength

Young's modulus* E is the slope of the stress–strain line. A large E means a small stretch at a given σ. A small E means a compliant member. E is a material constant, not a grade constant. Every structural steel, from S235 up to a quenched 42CrMo4, sits near E ≈ 210 GPa. Changing the grade does not stiffen the bar. It only raises the stress at which the line ceases to be a line.

σ = E × ε

What changes from one steel to the next is Re. S235 yields near 235 MPa; 42CrMo4 near 750 MPa. Under the same load the stretch is the same. The force at which the bar takes a permanent set is about three times higher on the alloy steel. Mixing the two numbers is the usual mistake on a drawing: E states how far the part moves, Re states when it yields.

A stiffer alloy is not thereby a stronger alloy, and a stronger alloy is not thereby a stiffer one. Aluminium 6082-T6 can match S235 on yield and still stretch about three times as much, because its E is about 70 GPa. Where the requirement is a precise centre-to-centre length, the section is sized on E. Where the bar must not take a permanent set, the section is sized on Re.

The strain printed here is the axial strain. The accompanying contraction of the cross-section is set by Poisson's ratio* and does not enter this check.

Reading the factor of safety

The result is given as s = Re / σ. One ratio, four quite different situations.

  • s < 1 — σ already above Re. Permanent set on the first application of the load. The area, the grade or the load has to change.
  • 1 ≤ s < 1.5 — the arithmetic holds. No spare capacity remains for a shock, a corrosion pit or a mill tolerance on thickness.
  • s = 1.5 to 2 — ordinary static work. Steel construction often adopts 1.5 on Re when the load is well known.
  • s = 3 to 5 — dynamics, impact, fatigue, or a failure with consequences for people. Lifting gear is required to sit at a factor of 5 on the breaking load of the cable, which is a different reference.

The ratio s is only as reliable as the Re and the A that entered it. A mill certificate for the actual heat is worth more than a catalogue Re. At a thread, the threaded section is the relevant area, not the nominal shank. A factor of 2 on a static hang is not a factor of 2 on a vibrating rod — that case belongs to a fatigue check, not to the elastic comparison on this page.

Worked checks on a tie, a cable and a specimen

  • M12 tie rod in S235 — tensile stress area* As = 84.3 mm². First yield at Re × As ≈ 19.8 kN, about two tonnes hanging. With s = 2 the load that stays inside that margin is one tonne, and only when the pull is centred and the thread is the section that was used.
  • Lift cable — the rule in view is s = 12 on breaking load. Pulleys, wear, reverse bending and the people underneath explain a margin that would look excessive on a static tie.
  • Tensile specimen — the test machine draws the σ–ε curve that this page reduces to two numbers: slope E in the elastic range, a plateau near Re, work hardening up to the ultimate tensile strength* Rm, then necking. The material list is that curve, flattened to E and Re.

A direct cross-check is Fe = Re × A, the force at first yield. When the applied F is a small fraction of Fe, the elongation is small and the factor of safety is large. When the two forces lie close together, the bar is already in the thin-margin band above, whatever units the printout uses.

After yield, from Re to the ultimate strength

Between Re and Rm the metal still carries load, but it flows. That continued rise is work hardening. Unloaded, the bar is longer than it was. Structural steels typically sit near Rm ≈ 1.5 × Re, with substantial elongation before separation. That reserve is why a mild-steel frame crumples rather than snaps, and why a shop can see the distress before the member parts.

Brittle materials — grey iron, glass, a ceramic insulator — have no such plateau. Sizing is against fracture, with larger factors, and Re is not treated as a yield point. The material list on this page is built for ductile metals and for wood along the grain. A cast-iron hanger in tension is the wrong use of the material; cast iron belongs in compression, where it is strong.

Once σ has passed Re, the elongation printed from the elastic formulas no longer describes the bar. Hooke's law has ceased to apply. What is required then is a plastic analysis, a test, or a larger section.

Steels, aluminium, titanium and timber

S235, S275 and S355 remain the economical choice for ordinary ties, and stepping up a grade is usually cheaper than machining a larger bar. Heat-treated C45 and 42CrMo4 raise Re sharply in the same outline; the cost is the heat treatment and a less willing chip. Stainless is a different bargain. A2 (304) covers indoor and food-adjacent service. A4 (316) adds molybdenum for chlorides — seawater, brine, a washdown. Neither grade matches a 10.9 bolt on raw Re; yield is not the reason those grades are selected.

Aluminium 6060, 6082, 2017A and 7075 sit at about one-third the density of steel and about one-third the modulus. At equal volume and equal stress the stretch is about three times that of steel, the same comparison already drawn for 6082-T6 against S235. At equal mass the area rises in proportion to the drop in density, so the axial stiffness EA stays comparable with steel rather than well above it. The practical reason for an aircraft longeron or a moving gantry is the load carried per kilogram on a high-yield grade, not a higher E.

Titanium Ti-6Al-4V pairs a high Re with a low density. Cost is what confines it to aerospace, implants and racing.

Wood along the grain is the exception in the list. It is weaker in tension than in compression — knots, slope of grain, a split that opens under pull. Structural steel carries the same stress either way. Copying a timber Re into a steel habit, or a steel Re into a timber habit, sizes the wrong failure.

Stress peaks at holes, threads and shoulders

σ = F / A is the average over the net section. At a hole, a fillet, a thread root or a weld toe, the local peak is the stress concentration factor* Kt times that average. Charts place Kt in the range 2 to 3 for everyday features; a sharp scratch or an unfused weld can sit higher. The bar does not part in the middle of the shank. It parts at the feature.

σpeak = Kt × (F / Anet)

On a static ductile bar a mild Kt can be accommodated: local yield, a little redistribution, the average stress remaining below Re. That redistribution ceases as soon as the load cycles. Fatigue responds to the peak, not to the average. A generous radius, a rolled thread, a polished blend — those are fatigue decisions, and they move the life more than another millimetre of diameter.

The calculation assumes a uniform stress: a straight member, a centred load, a constant section. A hole, a thread, a shoulder or an eccentric pull is not inside the printed figure. Parts fail at those details. Where the geometry is not a prismatic bar, Kt belongs in the check, or a local model does.

Net area at a thread, and at a line of holes

The field note beside the area is not decoration. On a threaded rod the load-bearing area is the tensile stress area As, a standardised value lying between the minor diameter and the pitch diameter — 84.3 mm² for an M12, not the 113 mm² of the 12 mm shank. Using the shank overstates the capacity by about a third. The trap is that many drawings write “M12” as though it were a 12 mm bar.

The same reduction appears on a drilled plate: Anet = t × (W − n d) for n holes of diameter d in a width W. The bolt can be adequate and the plate can still yield across the hole line. A row of holes in a flat bar is a tensile member with a reduced section, not a joint of some other kind.

ISO 898-1 bolts are rated on As. The shear page carries the same fork — shear plane through the shank or through the thread — and the same shortfall of about 25% when the wrong area is taken. The area entered is the area under tension.

Fatigue of a bar that cycles in tension

A static s = 2 is an acceptable margin for a hangar tie that never moves. It is not an acceptable margin for a rod that sees 106 cycles. For a polished, un-notched steel specimen the endurance neighbourhood lies around half of Rm; a thread, a corrosion pit or a weld toe brings that level down sharply. Aluminium has no true endurance limit — life is always finite — and 7075-T6 is particularly notch-sensitive.

The lift-cable factor of 12 is a fatigue-and-wear margin written as a static ratio. Reverse bending over a sheave, rain and a nicked wire are the reasons. The breaking load of the new rope is the reference, which avoids a separate Kt argument for every rope in a lift shaft. On a machine tie the peak is dealt with directly: a larger radius or a rolled thread, a lower live stress, or both.

This page does not return a life. It returns the mean stress. Where the load cycles, that mean goes into a fatigue calculation — Goodman, Gerber, or a code S–N curve — and Kt is treated as a primary input, not as a remark in the margin.

Square millimetres, centimetres and metres

Stress comes out in pascals when F is in newtons and A is in square metres. A 100 mm² bar is 1 × 10−4 m², not 100 m² and not 0.01 m². A missed conversion puts the stress out by a factor of 10,000 or of a million, which is how a 10 kN load on a small cross-section can appear to show that S235 is unbreakable. The unit picker on the form carries out that conversion, so the change of unit is not left to hand arithmetic.

A 10 kN pull on 100 mm² is 100 MPa. The same pull on 100 cm² — a 10 cm × 10 cm plate — is 1 MPa. The same pull on 100 m² is not a structural result. On the shop floor “a hundred square” means square millimetres. The calculator does not infer the unit. Square millimetres suit a bolt, square centimetres a small plate, and square metres only a member on the scale of a building.

Elongation has a related trap. L₀ enters the base calculation in metres, and ΔL is printed in millimetres. A strain of 0.001 on a 1 m bar is 1 mm of stretch. On a 50 mm gauge length the same strain is 0.05 mm, which a digital caliper can resolve and the eye cannot. When the elongation looks too small to matter, the length that was entered is the first quantity to recheck.

Glossary

Yield strength
The stress Re at which a pulled bar stops recovering its length and keeps a permanent elongation. Below Re the response is elastic. Above Re the elastic formulas on this page no longer describe the bar.
Hooke's law
In the elastic range, stress and strain stay in proportion: stress equals Young's modulus times strain. Twice the force produces twice the elongation, and removing the load returns the bar to its original length. The proportion ends at yield.
Factor of safety
The yield strength divided by the applied tensile stress, s = Re / σ. When s is greater than 1, the calculated stress is still below yield. How far s sits above 1 is the margin on that comparison.
Young's modulus
The slope E of the elastic stress–strain line. It states how much axial strain a given stress produces. It is a stiffness, not a strength: a high E limits the stretch and says nothing about the stress at which the bar yields.
Poisson's ratio
While a bar extends elastically, the cross-section contracts. Poisson's ratio is the lateral strain divided by the axial strain. The elongation calculated here is axial only; the lateral contraction is not evaluated.
Tensile stress area
The conventional area As used for the tension capacity of a screw thread. It lies between the minor-diameter area and the pitch-diameter area, and it is smaller than the plain shank. An M12 thread is rated on 84.3 mm², not on the 113 mm² of a 12 mm bar.
Ultimate tensile strength
The highest engineering stress on a tensile-test curve, Rm, reached after yield as the metal work-hardens, and followed by necking. For the structural steels discussed here it is typically about 1.5 times Re. The elastic calculation does not use it as a limit.
Stress concentration factor
Kt, the ratio of the local peak stress at a change of section — a hole, a fillet, a thread root, a weld toe — to the average stress on the net area. Everyday features often fall between 2 and 3. The uniform-stress formulas do not include this factor.

Related calculators